Few technologies captured the market's attention quite like artificial intelligence (AI) over the course of the last few years. It comes as no surprise, however, given the enormity of the technology's potential power -- at least according to its ambassadors. The Sam Altmans (CEO of OpenAI) and Dario Amodeis (CEO of Anthropic) of the world see a not-too-distant future in which their products transform our economy and fundamentally reshaped the way most firms do business.
But have you ever met a tech CEO that doesn't claim their company is changing the world? It's a good thing that AI optimism extends well beyond the C-suites of Silicon Valley. Investors can take comfort in the fact that organizations from the International Monetary Fund to the United Nations have made clear they believe AI will have a massive impact.
NASDAQ: NVDA
Key Data Points
Nvidia is riding the AI wave
The not-so-secret sauce powering this AI revolution are the uber-powerful chips the technology demands. When you ask ChatGPT how to make the best roast chicken (hint: mix a little baking powder and salt to get that skin extra crispy), a hive of graphics processing units (GPUs) spring to life. The system is big. Meta Platforms is building a data center in Louisiana, on which you could lay 70 football fields. As you can imagine, to fill such an area, companies need to purchase a lot of GPUs.
Without a doubt, no company has a hold on the AI GPU market like Nvidia (NVDA 1.56%) and it's been handsomely rewarded for it. Driven almost entirely by the sale of these chips, the company saw its revenue skyrocket since 2023. Check out the chart showing the sales growth compared to Meta (a major customer), Advanced Micro Devices (a rival chipmaker), and Alphabet (which relies on AI to improve its Google search engine, as well as in products such as Waymo, Nest, and YouTube).
NVDA Revenue (TTM) data by YCharts
Its growth has been monumental. In the same period, Nvidia also vastly expanded its margins while continuing to innovate to stay well ahead of the competition. The fact is, that no company has been able to touch it when it comes to producing the best chips around. And with the capacity to greatly outspend rivals in research and development (R&D) and the ability to attract and retain the best talent, it seems unlikely to me that this will change any time soon.
CUDA is key
While a lot of attention is paid to Nvidia's edge in raw power, perhaps its most important, enduring advantage is in its CUDA platform. Think of CUDA as a sort of language that allows programmers to use GPUs for purposes other than rendering graphics (for which they were originally created).
CUDA is the software foundation on which the complex scaffolding of AI is built. The higher-order development platforms and tools that almost all AI developers use are made to work with CUDA.
This means that if a customer wanted to switch to a rival's chips, they would need to invest enormous time and money transforming their entire workflows and workforces -- either through retraining or firing and hiring new developers. While this isn't impossible, the cost of this makes it a nonstarter for most. CUDA keeps Nvidia's clients locked in to its ecosystem, willing to pay a premium.
It's not an easy path forward
Despite Nvidia's dominance, there are still a lot of challenges that lay ahead. I don't see any major hurdles in the near future that Nvidia can't easily clear. All signs point to the company continuing its blockbuster sales over the next year. Its major big tech clients have all committed to record amounts of AI-focused capital expenditures, much of which is earmarked for data center expansion.
Longer term, however, a few big questions stand out. First, AI will have to begin delivering returns that can justify the vast sums being spent on its development. Promise and potential only get you so far. The shareholders of Alphabet, Meta, and their ilk will grow increasingly uneasy with the spending spree if a commensurate return isn't made clear. If this doesn't begin to happen within the next two years, I could see the money slowing.
Second, while Nvidia is likely to maintain its edge for the foreseeable future, tech is an industry defined by disruption. An upstart chip firm using radical new technologies could burst onto the scene. Or, perhaps in less dramatic fashion, its major clients develop their own chips and make the switch in spite of the CUDA moat I discussed.
Finally, Nvidia's success has become one of its biggest challenges -- at least for its stock price. The company delivered another monster earnings report last month, delivering nearly 80% sales growth year over year, yet Nvidia shares are down 21% in the last six weeks. There are macro factors that aren't helping, but the point remains that the market currently expects perfection from the company, and perfection doesn't exist -- unless you ask my fifth-grade violin teacher; then it does, and I'm just not capable of it.
Still, there are few companies with the vision that Nvidia has demonstrated. Nvidia will be able to adapt to the challenges that lie ahead. It won't be all roses by any means, but I also think that over time, expectations of perfection will fade, and Nvidia's bottom line will continue to grow substantially. And considering that after the stock's recent decline, its price-to-earnings ratio (P/E) is now the lowest in many years -- well before the AI craze took off -- I think Nvidia makes a smart investment.
